The sharp-eyed reader might have noticed that in last week’s article we mentioned that we’re crafting an integration tool called Tapestry. Our work on that tool uncovered what turned out to be a surprisingly rich subject: the dark art of programmatic error messages. Error messages in HRIS systems can reveal more about a product than any glossy feature page. Picture a manager approving a promotion, watching the spinner pause, then seeing a small message at the top of the screen: “Something went wrong,” the digital equivalent of “D’oh!” There are no explanations and no hint of what to do next.
Error handling is now recognized as a critical part of UX, especially as research has shown that users experience the product and its errors as a single whole. When a session times out and a lengthy performance review disappears, the person using the system does not separate “the good parts” of the system from “the glitch.” They simply decide whether the tool is dependable enough to use again.1 2 3 4 5
Current UX research treats error messages as communication that helps users recognize what went wrong, understand it and recover from it. That progression sounds obvious now, but it contrasts sharply with the old habit of serving up raw server codes or vague labels that read more like internal diagnostics. A confusing error can trigger support tickets, delay payments, or cause anxiety around sensitive data. 1 2 3 4 5 7
Modern UX writing and research reveal the architecture behind the best error messages. A useful message makes the problem clearly visible, using a combination of text, color, and placement that users can quickly notice, including those with different visual abilities. It states what happened in straightforward language, using familiar terms rather than internal system labels or numeric codes. It then offers a practical way forward, so the user is not left wondering whether to refresh, retry, contact support or go do something else. When the stakes involve money or employment, it often adds a note of reassurance, for example clarifying that payments were not sent or that new data entries remain safe. 1 2 3 4 5
The structure doesn’t need to be visible in the message because the user on the receiving end doesn’t care about the model behind it. They care that the message is understandable, reasonable and relevant to what they are trying to accomplish.
The contrast between old and new thinking is easy to see if we walk through a few common HRIS scenarios. A payroll administrator runs a cycle and sees “Error 500: Internal server error.” The label is technically accurate but useless as guidance. It doesn’t say whether any employees were paid, whether data was saved, whether it is safe to retry, or whether the admin should call someone in a panic. A hiring manager filling out a job requisition might submit a form and see “Invalid input. Try again.” They have no idea which field is at fault or what “invalid” means in this context, so they start guessing. Permission-heavy environments add another layer of frustration when messages like “Access denied” appear without context, leaving HR business partners unsure whether they have done something wrong or are encountering an expected restriction. 1 2 3 4 5
Unhelpful messages like these are cited often in analyses of poor error handling in a UX. They are opaque, sometimes feel punitive and place the burden of interpretation entirely on the user. Modern practice moves away from this. In the payroll use case, a more helpful message can acknowledge that the payroll run could not be completed due to a technical issue on the system side, clearly state that no payments were sent, and suggest trying again after a short delay or contacting payroll support if the issue persists. With that single change, the admin knows what happened, what did not happen and what to do next, which directly reduces stress and time wasted. 1 2 3 4 5
On the job requisition example from above, a generic “invalid input” can be replaced by a short, field-level explanation. Instead of a vague banner, the salary field might display a message explaining that the salary needs to fall within the approved range for the role, including the actual numbers. UX research shows the importance of placing these messages next to the field and using clear, concrete language that describes the requirement or constraint. In permission scenarios, a garden variety “Access denied” can be turned into a statement that describes what can’t be viewed and why, followed by instructions on how to request the relevant access from an administrator. This kind of wording respects the user’s role and provides a realistic path forward rather than a dead end. 1 2 3 4 5
Different categories of errors call for different patterns. Validation errors often involve typos, missing information, or values out of range in forms for employee records, job postings, or benefit elections. These issues are usually limited in scope, so the modern recommendation is to keep feedback local, specific, and as immediate as possible. The system can highlight only the fields that need attention, show an example of acceptable input, and avoid waiting until the entire form is submitted to reveal problems.9
Action failures belong to a different class. These can occur when a user attempts to approve a promotion, export a large report, import data or update many records in bulk. Here the user has already invested time and effort, and the consequences of system failure are more frustrating. Contemporary UX guidance encourages systems to preserve work rather than discarding it, to name the action that failed in clear language, and to explain what adjustments might lead to success. An HRIS that cannot export a report due to file size constraints can say so plainly, recommend filtering to a shorter period or smaller population, and invite the user to try again rather than leaving them with a generic error label. 1 2 3 4 5
System outages and service disruptions are another common situation. In HR, a banner that vanishes or a vague “service unavailable” message near a benefits deadline understandably creates concern (to say the least). Good UX guidance urges teams to own such issues clearly as system problems, to describe the impact and to set realistic expectations about timing when possible. A message that explains that benefits enrollment cannot be displayed due to a system outage, confirms that existing selections remain saved and suggests coming back later that day or contacting HR if the deadline is immediate gives users enough information to decide what to do next. Studies of error states show that this kind of transparency will reduce confusion significantly and unnecessary support contact. 1 2 3 4 5
Security and permission-related messages need particular care. Strong access controls are necessary in HR, but the way they are communicated can either protect privacy gracefully or leave people feeling accused. Modern UX writing recommendations encourage neutral phrasing in these situations, explaining that a person does not have permission to view a certain type of data and pointing them toward the correct administrative contact if access is genuinely needed. The emphasis is less on scolding and more on simply explaining the rules of the system.11
Tone matters throughout. Error messages in HR contexts carry more emotional weight than those in many consumer apps. A glitch in a Netflix queue is annoying but a glitch in a pay calculation or performance rating can have real, deleterious effects about how someone feels about their job. Recent writing on error communication calls for plain, respectful language that avoids jokes or playful phrasing when money, privacy, or employment are involved. It also warns against wording that explicitly blames the user. Rather than telling someone they “did not complete the required fields,” the system can state that “some required fields are missing” and invite them to review highlighted sections. This subtle shift feels more collaborative to the user than fault-finding. 1 2 3 4 5
Respect for user effort is another recurring theme. Designers and researchers stress the importance of preserving typed content when something goes wrong, retaining filter settings when an export fails, and consolidating multiple validation issues in a way that is easy to scan. Work that focuses on error states often recommends tracking metrics such as recovery time and completion rates, then using those numbers to revisit how errors are handled. In HRIS environments, where flows are often long and complex, these practices can noticeably affect how people perceive the reliability of the system.7
Contemporary UX guidance suggests treating error handling as a design topic in its own right. Teams can benefit from an internal style guide that spells out tone, common structures, and examples tailored to the product. That might include templates for session timeouts during performance reviews, messages for data import failures, language for effective-date conflicts in position changes and standard phrasing around compensation visibility. Some organizations borrow or adapt scoring rubrics to review messages against criteria such as clarity, helpfulness, and tone, then refine the weak spots. 1 2 3 4 5
Early in the design process, this way of thinking shows up in wireframes that include error states alongside the “happy path.” Instead of a single mockup of a successful promotion approval, designers and writers explore what happens when network connectivity fails, when permissions are missing, or when effective dates conflict. Later, teams study these flows to see where users encounter errors, how quickly they recover, and where they abandon tasks altogether. Those insights will reveal a small set of recurring issues that are worth addressing in both UX and system behavior.9
Error messages in complex systems are moving away from generic shrug-like responses and toward specific, context-aware communication that treats users as partners in resolving problems. Generic “Something went wrong” text is being replaced with messages tied to the user’s action, the relevant data, the likely cause and often suggesting a solution. Blaming language is being replaced by neutral explanations, and one-size-fits-all warnings are giving way to patterns that recognize the difference between a simple typo and a failed payroll run. Teams increasingly treat these states as measurable parts of the product, tracking them alongside primary usability metrics and adjusting their design as they learn from real usage.8
Thoughtful error messages influence how quickly managers and employees complete tasks, how often they seek help and how much confidence they feel in a system that handles sensitive aspects of their work life. Because many improvements involve wording, placement and modest interaction changes rather than deep technical rewrites, they can often be introduced incrementally, starting with the most visible and painful flows. 1 2 3 4 5
A message that appears at the top of the screen during a stressful moment will never be the star feature of an HR platform. But to users, though, it embodies the character of the product. When it is clear, relevant, and actionable, it supports the relationship between people and the systems they depend on. When it is vague, blaming, or opaque, it weakens that relationship. The evolution of UX thinking around error messages invites teams to treat these small pieces of text as though they’re designing in trust in the system instead of treating them as afterthoughts. 1 2 3 4 5
Sources:
- Nielsen Norman Group. “Error-Message Guidelines.” 2023.
https://www.nngroup.com/articles/error-message-guidelines/ - Nielsen Norman Group. “Error Messages: 4 Guidelines for Effective Communication (Video).” 2024.
https://www.nngroup.com/videos/error-message-communication-guidelines/ - Nielsen Norman Group. “Error Messages 101 (Video).” 2024.
https://www.nngroup.com/videos/error-messages-101/ - Nielsen Norman Group. “Create Efficient Error Messages (Video).” 2024.
https://www.nngroup.com/videos/efficient-error-messages/ - Nielsen Norman Group. “Hostile Patterns in Error Messages.” 2022.
https://www.nngroup.com/articles/hostile-error-messages/ - Nielsen Norman Group. “An Error Messages Scoring Rubric.” 2023.
https://www.nngroup.com/articles/error-messages-scoring-rubric/ - Smashing Magazine. “Designing Better Error Messages UX.” 2022.
https://www.smashingmagazine.com/2022/08/error-messages-ux-design/ - UX Content Collective. “How to write error messages.” 2025.
https://uxcontent.com/how-to-write-error-messages/ - Interaction Design Foundation. “Error Messages UX with Vitaly Friedman (Master Class).” 2024.
https://ixdf.org/master-classes/how-to-design-better-error-messages-ux - NN/g. “How to Increase the Visibility of Error Messages (Video).” 2024.
https://www.nngroup.com/videos/error-message-visibility/ - YouTube. “UX TUTORIAL: Designing an error message the right way.” 2025.
https://www.youtube.com/watch?v=dUYW7vEWDWg
